Close Window

Next-Generation SAR Amazon Mosaic: JERS and RADARSAT

Leland E Pierce, Univ of Michigan - EECS Dept., lep@umich.edu (Presenting)

We discuss the development of the next-generation SAR Amazon mosaic using both L-band JERS and C-band RADARSAT data. Rather than using rubber-sheeting-based techniques, this project uses custom orthorectification programs to allow for precise positioning, geometric correction, and radiometric correction made possible by the recent 90-meter-spacing digital elevation model provided by the Shuttle Radar Topography Mission that flew on the Shuttle in 2000. This topographic data allows for the correction of many SAR-related image distortions. We rely on accurate satellite orbital information as gathered by NORAD and published on the web to correct for the usually quite inaccurate orbital information that comes with each image. Since we are building upon work done by JPL when they processed the original Amazon mosaic using JERS (L-band) SAR data only, we are using the same 34 tiles that they used to divide up the Amazon into manageable pieces. Each original JERS or RADARSAT image is then orthorectified onto a tile, and subsequent images are then allowed to overwrite this same tile until a complete tile is done. At this point the SAR data is relatively seamless due to the orthorectification process. Geometric accuracy will be assessed using natural features such as rivers and coastlines. The resulting data will be made available to the public.

Presentation Type:  Poster

Abstract ID: 19

Close Window